Data Preparation

Model your data as a series of process observations or measures that are associated with an outcome of interest.  Compose each observation (row) as a common set of features (aka., independent variables or factors).  Both features and outcomes are either numerical (age, HbA1c, BP, etc.), binary (yes/no, true/false, etc.) or categorical (Gender, Service line , Floor unit, Shift, DRG, etc.).

Provide a CSV file with a header row labeling each column.

  1. The first column is an alphanumeric index starting with a letter that is used to uniquely identify the data row.
  2. The second column is an alphnumeric label for the group of rows the data belongs to.  Use this if you want to look for anomalies within groups of rows as opposed to across all rows.   If there are no groupings then just label the all values in the 'GROUP' column the same.
  3. Provide additional columns of either categorical data or numeric data.
  4. The first row is a user supplied alphanumeric header for each column

 Sample Table:

INDEX GROUP COL2 COL3 COL_n
ROW_1 Group_A Red Hot Early
ROW_2 Group_A Yellow Cold Late
ROW_n Group_A Yellow Cold On-Time

Learn more »

Submission

Try it now with your data.

Pattern Discovery and Prediction

  1. We engineer features, as needed, to remove noise and improve the quality of findings and predictive models.

  2. Using your domain knowledge we select the data columns of interest for processing

  3. We auto-tune the machine learning to your data so that turnaround times can be exceedingly fast.

Emailed Results Provide Early Insight

  1. You receive a list of the top anomalies rated by confidence

  2. You are provided with a chart that profiles factors associated with the top anomalies.  This allows for generation of simple process decisioning rules fostering implementation of realtime corrective process interventions.  

  3. We bin continuous data into 1-dimensional clusters before processing to enhance explainability in discovered findings.  As an example, we would priovide a range of temperatures associated with anomalous results.

News

HTO Magazine